The general aim of this project is to further the analysis of aphasic syndromes in terms of the processing models that have emerged from contemporary work in psycholinguistics. The work that is planned for the next funding period focuses on the following areas: (1) Sentence processing in agrammatism. Further efforts will be made to identify the factors underlying the sentence comprehension deficits of agrammatic aphasics. Approaches will include systematic modification of structures that are well comprehended by these patients to produce structures that are poorly comprehended. (2) Studies of agrammatic production. Procedures will be developed for the quantitative assessment of agrammatic speech samples. In addition to providing a means of identifying agrammatic subjects, this metric will be used to track the recovery process in agrammatism and to compare agrammatic production patterns with those of other aphasics. Controlled studies of sentence production will also be performed to further specify the nature of the production impairment in agrammatism. (3) Studies of lexical impairment. An attempt will be made to subdivide anomic aphasics into two classes reflecting impairment at different stages (semantic vs. phonological) of the word retrieval process. These distinctions will be exploited in studies of speech production. (4) Speech errors analyses in fluent aphasics. Error patterns in extended speech samples and structured tasks will be analyzed within the framework of stage models of speech production. These studies should contribute to the functional descriptions that are required for the delineation of brain-language relationships, and to the development of more principled approaches to patient evaluation and rehabilitation.